TAN Yinliang, Professor of Decision Sciences and Management Information Systems at China Europe International Business School: How Does AI Rewrite the Rules of Productivity? | 36Kr 2025 AI Partner Conference for All Industries

On August 27th, the 2025 AI Partner Conference for All Industries, jointly hosted by 36Kr and China Europe International Business School (CEIBS), grandly kicked off at Zhongguancun Software Park in Beijing. The conference, themed "Chinese Solutions," is divided into two major chapters: "Chinese Solutions" and "Who Will Define the Next AI Era." It focuses on four major topics, including "The Golden Age of Chinese Innovation," "Can Superintelligent Agents Become the Core Form of Next-Generation AI?" "Chinese Solutions Reshape the Global Technological Competition Landscape," and "The Prosperous Scene of the Integrated Innovation of AI and All Industries," comprehensively presenting the latest breakthroughs and ecological systems of Chinese AI, sharing the growth path and future prospects of Chinese-style AI, and exploring the innovative models of Chinese solutions.

On the same day, Tan Yinliang, a professor of Decision Sciences and Management Information Systems at CEIBS and an expert in AI applications and industries, delivered a keynote speech titled "How AI Drives Business Value and Productivity Growth."

The following is the content of the speech, edited by 36Kr:

Respected Secretary Ma, Teacher Xiao, and General Manager Feng, dear guests:

Good morning, everyone!

I'm very honored to be the first one to stand here today and share with you the topic of "How AI Drives Business Value and Productivity Growth." Yesterday, the State Council just released the "AI +" Action Plan, which makes it even more necessary for us to think about: How should we understand artificial intelligence in the next decade? And how will this technology change our economy and society?

Last year, Dr. Wu Wenda, a former AI scientist at Baidu and a world authority on artificial intelligence, made a wonderful analogy at an event: "AI is the electricity of the new era." Just as electricity completely changed human society more than a hundred years ago, it's hard for us to imagine which industry will be outside the influence of AI in the next few years. That's why we'll spend a whole day today discussing how AI penetrates all industries, which is highly consistent with the theme of this conference.

To understand how AI can improve productivity, perhaps we can find inspiration from the history of the electrical revolution. Let's first review the development of the electrical revolution: In 1831, Michael Faraday discovered the phenomenon of electromagnetic induction, laying the foundation for the application of electricity; in 1879, Thomas Edison invented the light bulb, lighting up the human night ever since; in 1880, Nikola Tesla and George Westinghouse jointly developed the alternating current transmission technology, solving the problem of long-distance power transmission; in the early 20th century, electric motors gradually became popular. Compared with steam engines, they are more efficient, flexible, and easier to control.

It's worth noting that the light bulb was invented in 1879, and power stations were built in New York and London in 1881. Technically, the electrical era seemed to have begun in the 1880s. But if we could travel through time, we'd find that there was almost no economic evidence in the early 20th century to prove that the electrical revolution had improved productivity.

This leads to a profound question: Electricity is so important, a technology we can't live without today. Why did it take more than 30 years to truly change productivity? American economist Paul David from Stanford University gave the answer in a 1990 academic paper. To help you better understand, I used AI to generate a picture of a factory in the steam engine era: At that time, factories had a centralized layout, with a huge steam engine in the center, transmitting energy through belts and conveyors. All processes and workstations had to be as close to the steam engine as possible.

When the electrical technology first emerged, people's thinking was straightforward - replace the huge steam engine with a huge electric motor just for "cost reduction and efficiency improvement." But such a simple replacement didn't bring obvious improvement to productivity. It wasn't until more than 30 years later that people gradually realized that they needed to change management thinking, transform organizational structures, reshape corporate cultures, and even redesign factory layouts to truly improve productivity.

If we use one word to summarize the key to electricity's more than 30 - year journey to change productivity, it's "management." Specifically, people made these changes: First, they popularized the unit drive system, disassembled the huge electric motor into multiple small ones, and redesigned the factory layout accordingly, making the flow of goods smoother and more efficient, while also improving the factory environment and reducing noise and pollution; second, they promoted the complementary development of new technologies, such as combining automated production equipment with electricity to achieve 24 - hour continuous production, and delegated power to industrial workers, redesigning contracts and incentive mechanisms.

These changes finally brought remarkable results: In the 1910s, more than 30 years after the invention of the light bulb, productivity in the United States soared. In 1913, Ford's annual car production reached 250,000 units; the productivity growth rate of the US manufacturing industry exceeded 5%, laying a solid industrial foundation for it to become the most powerful economy in the 20th century.

Looking back at the history of the electrical revolution, now that we're at the starting point of the AI era, how should we embrace this technology? I've summarized three insights:

First, simple replacement often doesn't work. Just like installing a steam engine on a traditional sailing ship without modifying the ship's structure will only make it slower. If we only see AI as a substitute for existing technologies and only focus on "cost reduction and efficiency improvement" instead of creating new value, it's hard to change the essence of business activities.

Second, it takes a long time for a major technology to mature, penetrate all industries, and release productivity. Behind this is the "Productivity J - Curve" theory: After introducing an important technology, productivity usually declines slowly at first and then gradually rises. Because any key technology needs time to adapt and integrate to truly affect economic activities.

Third, comprehensive innovation is the key to success. Technological transformation not only requires new technologies themselves but also comprehensive innovation and transformation of business processes, business models, organizational structures, and corporate cultures.

By comparing the development stages of the electrical era, we can see more clearly where AI stands today:

The first stage is the "early stage of technological breakthrough." In this stage of the electrical era, people were excited about the popularization of electric lighting, but they didn't modify factory layouts and processes, and industrial productivity hardly improved. In 2022 and earlier, AI also went through a similar stage - tools like ChatGPT and Copilot were used for office assistance, with limited application scope and still in the exploratory period.

The second stage is the "early application period." In the electrical era, new factories and specific processes began to use electric motors, and efficiency improved but not significantly; in the AI era, the technology is mostly used to improve individual efficiency locally, and the effects are hard to show in financial reports. A study by MIT shows that 95% of US companies said AI had no impact on them; a report by McKinsey in July this year also pointed out that 80% of companies had deployed AI systems, but the same proportion of companies said it had no substantial impact on their operations. The root cause is that they're still in this stage.

Next, we'll enter the "structural transformation period" - this is the most critical stage for Chinese enterprises in the next 3 - 5 years and the moment when AI moves from the "productivity trough" to the inflection point. In the electrical era, people needed to restructure factory layouts, train workers, and adjust supply chains; in the AI era, we need to think about how to transform business processes (existing SOPs are designed for humans, not AI) and restructure systems. This stage requires high short - term investment and has a long return cycle, but the long - term returns are huge.

Finally, we'll enter the "mature expansion period," which is the season for reaping the value of technology. For AI, I'm particularly optimistic about vertical intelligent agents - they can be deeply embedded in the core of various businesses, realizing process automation and intelligent decision - making, and then giving rise to new business models and competitive advantages.

The key elements for success in the AI era lie in strategic focus, business process reengineering, upgrading of technology and data frameworks, and improvement of governance systems. We'll detail the practical methods for these in CEIBS courses.

Finally, I'd like to share a quote from Demis Hassabis, a Nobel laureate in chemistry: "Our goal is to solve the problem of intelligence and then use the intelligence created by humans to solve all other problems."

Thank you!